3.

Trade-Labour Link: A Post-Seattle

Analysis*

Arvind Panagariya
___________________________________________________

1. INTRODUCTION

The first Ministerial Conference of the World Trade Organization (WTO),

held in Singapore during 9-13 December 1996, had a rough start. The United
States (US) had wanted to create a WTO working group on labour standards,
which developing countries bitterly opposed. In the end, developing countries
prevailed and the US effectively agreed to delegate the subject to the
International Labour Organization (ILO). In the only paragraph on the
subject, the Singapore Ministerial Declaration stated:

We renew our commitment to the observance of internationally recognized

core labor standards. The International Labour Organization (ILO) is the
competent body to set and deal with these standards, and we affirm our
support for its work in promoting them. We believe that economic growth and
development fostered by increased trade and further trade liberalization
contribute to the promotion of these standards. We reject the use of labor
standards for protectionist purposes, and agree that the comparative advantage
of countries, particularly low-wage developing countries, must in no way be
put into question. In this regard, we note that the WTO and ILO Secretariats
will continue their existing collaboration.

Most readers, including legal experts, will agree that this statement
conclusively designates the ILO as the competent body to 'set and deal' with
labour standards and leaves little room for the WTO to take lead. So as not to
leave any doubt about the validity of this interpretation, Trade and Industry
Minister of Singapore Mr. Yeo Cheow Tong, who also chaired the
Ministerial Conference, went on to note in his concluding speech, 'Some
delegations had expressed the concern that this text may lead the WTO to
acquire a competence to undertake further work in the relationship between

5
trade and core labour standards. I want to assure these delegations that this
text will not permit such a development'.
For two years following the Singapore Ministerial, no WTO Member
including the US questioned this understanding and interpretation of the
Singapore Declaration. But beginning in January 1999, the US began to place
proposals before the WTO for the establishment of a work program in the
WTO that would address trade issues relating to labour questions.1 Later, as
the preparations for the Seattle Ministerial got under way, it became
increasingly clear that the US would return to its original demand for a
working party on labour standards in the WTO. By 13 October 1999, the
Financial Times had reported that the US was set to call for such a working
party. In early November, the US did table a formal proposal at the WTO. 2
The opposition to a WTO working party on labour standards by the major
developing countries such as Brazil, Egypt, India and Malaysia, had been
well known. At Singapore, these countries had fought hard to have the matter
delegated to the ILO and had even given up their opposition to a working
party on investment in return. Given the intellectual merit of the case, they
had also been supported in their endeavour by some key developed countries,
most notably Australia and Great Britain but also Germany and Switzerland.
Therefore, the decision by the US in early November to go back on the letter
and spirit of the Singapore Declaration immediately cast a shadow of doubt
on the success of the Seattle Ministerial.3
Initially, the US attempted to soften the impact of its demand by arguing
that the sole purpose of the WTO working party would be to generate
discussion on how labour standards impact trade and development. But later,
while the Seattle Ministerial Conference was in progress, President Clinton
went on to tell the Seattle Post-Intelligencer that he would like the working
party to define core labour standards, which should then be incorporated into
all trade agreements and subject to trade sanctions. This explicit
acknowledgement by the US of its ultimate goal behind the demand for the
working party instantly united the major developing countries in their
opposition to the US proposal and sealed the fate of the Seattle conference.

1. Two such proposals were made, one in January and the other in July 1999. See the document
'US Proposal on Labor Rights' in Jenkins (1999).
2. See Financial Times (1999).
3. For example, in a letter published in the Financial Times, 10 November 1999, I had already
stated that 'if the US persists in its demand, developing countries, which are bound to bear the
brunt of the link [between trade and labour standards], will be fully justified in walking away
from the Seattle talks'. See Panagariya (1999a). In Panagariya (1999b), I offered a
comprehensive discussion of the agenda for a possible Seattle Round and argued that
developing countries should make every effort to limit the agenda to trade liberalisation and
the Uruguay Round built-in agenda.

6
 The demand for a WTO working party on labour standards by the US,
though unsuccessful in Seattle, has returned the contentious issue of the link
between trade and labour standards to the center stage of multilateral trade
negotiations. Motivated by this fact, I undertake a detailed and systematic
analysis of the subject in this chapter. I begin first by dissecting the
intellectual case for the link in section 2. In section 3, I describe the so-called
'core' ILO Conventions. In section 4, I identify the sources of pressures for
higher labour standards that led to the failure in Seattle as also the pressure
being exerted within the current system on some developing countries to
raise labour standards to preserve their privileges under the Generalized
System of Preferences (GSP). Finally, in section 5, I conclude the chapter
with suggestions for future progress towards the resolution of the current
impasse.

2. EVALUATING THE CASE FOR LINKING TRADE AND

LABOUR STANDARDS
The case for labour standards was originally subject to a critical examination
by Bhagwati (1995, 1996). Here I offer an updated discussion of the subject,
though it may be acknowledged at the outset that Bhagwati’s original
arguments and critique remain valid today. The advocates of the linkage have
offered few new arguments that would seem persuasive.

2.1 The Efficiency Issue

With respect to efficiency, two simple analytic points may be made. First, in
general, optimal labour standards are not uniform over time or across
countries, either from the national or global welfare standpoint. The changes
in marginal benefits and costs of labour standards as, for example, due to
changes in income or productivity in 'supplying' labour standards, cause
optimal labour standards to vary over time as well as across nations.
Therefore, there is no presumption that the differences in observed labour
standards between two countries necessarily imply deviations from optimal
standards in at least one of them. Nor is there a case for the harmonisation of
labour standards internationally (or over time) on the ground that it promotes
efficiency at the global or national level. Equally, the observed labour
standards can be below the optimal level both in developed and developing
countries; just because labour standards are higher in developed countries
does not mean that they are optimal.
Second, the targeting literature, pioneered by Bhagwati and Ramaswami
(1963) and Bhagwati (1971), tells us that when an economy is in a sub-
optimal equilibrium, the first best policy is to correct the underlying

7
distortion at its source. Once this is done, there is no need to intervene
elsewhere in the economy. Thus, if the market happens to produce sub-
optimal labour standards, we should correct this distortion directly rather than
through an indirect instrument such as trade sanctions. Under the direct
approach, once labour standards have been set at the optimal level, free trade
remains the optimal trade policy in the traditional sense. Purely from an
efficiency standpoint, a case cannot be made for linking trade and labour
standards.

2.2 Evaluating the Appropriateness of Pursuing Labour Standards

in the WTO

As argued above, if labour standards were set at their optimal levels

everywhere, the discussion of raising them through the WTO instrumentality
will be moot. Instead, we will only need for the WTO to promote further
trade liberalisation. To proceed, therefore, we must assume that the existing
labour standards are sub-optimal, possibly in developing as well as developed
countries. The question then is whether the WTO is the right institution to
achieve this goal. Subsumed within this question, of course, is the issue of
linkage itself since the key feature that distinguishes the WTO from the
alternative institution, the ILO, is its ability to back up agreements on labour
standards by trade sanctions. But in this sub-section, I will only touch on this
issue, leaving its detailed discussion to the following sub-section.
There are at least three criteria on which we can judge the appropriateness
of the inclusion of areas such as labour standards, intellectual property rights,
competition policy or investment policy into the WTO. First, is the area
sufficiently closely related to trade? That is to say, do countries choose the
policies in that area principally to influence trade or to fulfil other objectives?
Second, will the inclusion of the subject improve world welfare? And third,
will the inclusion improve the welfare of each WTO Member? The second
criterion is, of course, necessarily fulfilled if the third one is. As such, the
second criterion is weaker and is not likely to be accepted by countries that
stand to lose, unless the beneficiaries of the discipline compensate them.
It is immediate that trade policy, which the WTO has been designed to
oversee, meets all these criteria. The principal objective behind trade barriers
is to restrict trade flows.4 Trade liberalisation is beneficial to the world and it

4. Interestingly, when the principal objective is viewed to be different from protection, the WTO
often accommodates trade restrictions. Thus, it permits temporary quantitative restrictions
when the objective is to overcome temporary balance of payments difficulties. Likewise, in
the past, when economic thinking (now largely discredited) admitted a role for trade
restrictions to achieve the development objective, the General Agreement on Tariffs and

8
is also beneficial to individual countries that engage in reciprocal bargains
under the auspices of the WTO. Trade liberalisation is a win-win activity.
By contrast, the trade-labour link fails to meet any of these criteria with a
reasonable degree of confidence. Rarely do countries set labour standards to
influence trade flows. There is no doubt that these standards have an
influence on trade flows but no one will argue that this is reason enough to
subject them to the WTO discipline. After all, exports equal domestic output
minus domestic consumption so that any change affecting either of these
quantities must affect trade unless the good happens to be non-tradeable!
Countries choose labour standards based on the prevailing socio-politico-
economic conditions. In developing countries such as India, child labour
existed long before trade acquired any significance at all. Likewise, much of
the labour legislation in the country was enacted to fulfil the perceived needs
of labour rather than to fulfil the needs of the industry to be competitive vis-
à-vis foreign sources either at home or abroad. Quite the contrary, the
prevailing political economy led to such high labour standards in some
dimensions that many domestic firms could only survive behind a high
protective wall.5 In the view of many economists, today, these labour laws
constitute a barrier to successful trade liberalisation. In effect, we have a case
of labour standards driving trade policy rather than the other way around.
Turning to the other two criteria, the link between trade and labour is not
a win-win policy and may very well lead to a decline in world welfare.
Chances are that the trade-sanctions bullet will miss its target. Consider child
labour. One possibility is that countries will fail to meet the WTO standards
leading to the imposition of trade sanctions. If so, no improvement in labour
standards will be achieved and the gains from trade will be reduced. World
welfare will be necessarily reduced. Alternatively, child workers may simply
be shifted from producing exports into alternative activities. Again, there will
be no net reduction in the aggregate volume of child labour while the wages
received by the children in alternative employments will be lower or the
working conditions worse. Once again, world welfare will be reduced.
What the trade-labour link tries to accomplish is to kill two birds with one
stone; use the WTO to achieve both free trade and higher labour standards. In
technical terms, the link seeks to hit two targets with one instrument. But as
the first Nobel laureate in economics Jan Tinbergen demonstrated, in order to
be successful, one would normally require at least as many instruments as

Trade (GATT), the predecessor institution of the WTO, readily permitted permanent
quantitative restrictions in developing countries.
5. In the organised sector, workers are paid wages that are substantially higher than elsewhere in
the economy, and cannot be fired. This means that even when a firm becomes unprofitable, it
cannot exit. Often such firms are declared 'sick', with the government taking charge of them
and paying the high wages at the taxpayer's expense.

9
he/she would targets. Taking advantage of this insight, a recent statement
entitled 'Third-World Intellectuals and NGOs: Statement Against Linkage'
(TWIN-SAL) argued that the best course to promote labour standards is to
pursue them through an alternative institution, the ILO, and leave the WTO
the task to promote free trade. This is also consistent with the Singapore
Declaration.

2.3 Preserving One’s Moral Values

The trade-labour link effectively requires countries to raise standards to a pre-

specified level or face trade sanctions from other countries. One argument
offered in defence of this policy is that a country that adheres to higher labour
standards within its national boundaries has the moral right to suspend trade
with another country that does not adhere to equally high labour standards.
For instance, if the US subscribes to values that do not admit child labour and
has itself outlawed the practice, it should also have the right to suspend
imports made by child labour in other countries. Why should US citizens
have to compromise their values to accommodate the imports from abroad?
There are two problems with this line of reasoning. First, when the US
chooses to outlaw child labour from its territory, it also chooses to pay the
cost of the change by foregoing the output produced by potential child
workers and by committing resources to educational facilities for them. But
when it asks other countries to also abandon the practice because it will help
promote a value held dear by its own citizenry, the cost, in terms of the
output foregone and additional resources spent on education, is borne by the
trading partners. These costs are not trivial. According to a study undertaken
by Consumer Utility and Trust Company (CUTS), a NGO based in India, it
will cost anywhere between US$12 billion to US$18 billion per annum in
India alone to send all existing child workers to schools. It is unlikely that the
US and other developed countries would be willing to bear even a tiny
fraction of this cost.
Second, if a common set of labour standards is to be adopted at the WTO
to promote higher moral values, Member countries must share these values
with sufficient conviction that they are voluntarily willing to subject
themselves to trade sanctions as an instrument of enforcement. But there are
few moral values that are shared universally with such conviction.
Proponents of the trade-labour link often give the impression that there is a
general agreement on the so-called 'core' labour standards among the WTO
Members. To substantiate the argument, they refer to the ILO Conventions

10
on 'core' labour standards that many countries have ratified. 6 Quite apart
from the fact that the ratification may simply be a feel-good act or an
expression of philosophical belief in the values embedded in the
Conventions, the latter are quite far from being universally ratified. For
example, the US itself has ratified only one of the seven traditional 'core' ILO
Conventions.7 Only 73 out of 175 ILO members have ratified all seven
Conventions. At 101 countries, the Minimum Age Convention remains the
least ratified. Thus, even in the absence of trade sanctions, countries show
very different preferences with respect to what are supposedly 'universally
agreed' standards.
Recently, Rodrik (1997) has questioned the logical consistency of the
position taken by the critics of the 'preservation of moral values' argument.
To explain his argument, consider a US corporation, which must find a
cheaper source of supply of some of its labour-intensive components if it is to
survive. Suppose further that it has two options. Under the first option, it can
outsource the components to a local Honduran firm that runs a sweatshop
operation. Under the second option, it can open a domestic sweatshop at the
Mexican border and employ the same Hondurans as migrant workers. The
costs of production, the wages received by Hondurans workers (net of
migrations costs) and working conditions are identical in both cases.
Therefore, if one approves of one option, he or she necessarily approves of
the other option. Rodrik notes, however, that this is not so:

Interestingly, the vast majority of the economists who have no difficulty with
the outsourcing example would also accept that it is not good public policy to
relax labor standards for migrant workers to the point of allowing sweatshop
conditions. Clearly, there is an inconsistency between these two positions.
There seems to be a greater coherence in the behaviour of the lay public,
which reacts to with equal outrage to the two versions of the parable -
outsourcing versus migration - than in the perception of the economists.8

6. Though a more detailed discussion of the ILO Conventions appears in Section 3, it may be
noted here that traditionally seven conventions are included among the ‘core’ conventions:
Forced Labour Convention, 1930 (No. 29), Abolition of Forced Labour Convention 1957
(No. 105), Freedom of Association and Protection of the Right to Organize Convention, 1948
(No. 87), Right to Organize and Collective Bargaining Convention, 1949 (No. 98), Equal
Remuneration Convention, 1951 (No. 100), Discrimination (Employment and Occupation)
Convention, 1958 (No. 111) and Minimum Age Convention, 1973 (No. 138). Of these, the
US has only signed the Convention on Abolition of Forced Labour, 1957 (No.105). In 1999,
a new convention, called the Convention on the Worst Froms of Child Labor (C182) was
signed. The ILO now lists this convention along with the other seven as well and,
presumably, includes it among the ‘core’ standards.
7
These do not include the Convention on the Worst Forms of Child Labor (C182) mentioned in
the previous footnote. This convention had been ratified by 41 countries at the time of
writing.
7. Rodrik (1997), p. 34

11
 On the surface, this appears to be a compelling critique of the position
taken by the opponents of trade sanctions against countries with low labour
standards. Yet, upon closer examination, contrary to Rodrik’s contention, it is
the position of the 'lay public' that is logically inconsistent and the view of
the 'vast majority of the economists' that is consistent. Thus, when the 'lay
public' shows outrage against the poor treatment of migrant workers, it wants
them to be treated on par with US workers with the cost of such treatment
falling on the corporation and hence the US economy. But when the 'lay
public' shows outrage against sweatshops in Honduras, it wants trade
sanctions that place the burden of upholding their moral values on the
Hondurans! Logical consistency would require that the 'lay public' be willing
to offer the Honduran government the cost of bringing the working
conditions in Honduras to the US level as well.
Likewise, the apparent contradiction in the position of the 'vast majority of
economists' is resolved once we recognise that when foreigners come in the
midst of post-Renaissance societies, especially the US, the local population
begins to view them as one of their own. The willingness to confer the same
rights on migrant workers as those available to local workers is an outcome
of this empathy. Moreover, the acceptance of sweatshops abroad is a vote
not against the rights and well-being of the workers employed therein, but
against trade sanctions that will bring an even worse alternative. Thus, there
is no contradiction between the advocacy of equal rights for migrant workers
and the opposition to trade sanctions when the same workers are employed in
their own country under sweatshop conditions.
Indeed, the position attributed by Rodrik to the 'lay public' appears more
coherent and understandable when considered from the viewpoint of the US
labour groups. Within the framework of his parable, weaker labour standards
in Honduras mean greater competition through trade in labour-intensive
industries in the US. Likewise, permitting sweatshop conditions for migrant
labour on the US soil gives greater incentive to US firms to employ the latter.
Both policies put pressure on the wages paid to local workers.

2.4 The Fair Trade Issue

Yet another argument offered by the proponents of the trade-labour link is

that lower labour standards in developing countries give them 'unfair'
competitive advantage over their developed country counterparts. Deep
down, this is essentially the age-old 'pauper-labour' argument that labour
unions have repeatedly used to seek protection for labour-intensive industries
in developed countries. Traditionally, the argument has relied primarily on
the existence of low wages in labour-abundant countries as the source of

12
'unfair' advantage. In its current incarnation, the reach of the argument has
been widened by including a whole host of labour standards among the
sources of unfair advantage.
As has been pointed out repeatedly by economists in textbooks and op-ed
articles alike, this argument is in direct conflict with the basic principle of
comparative advantage. Virtually every textbook on international trade
describes the pauper-labour argument as a common fallacy. The simple point
is that high wage countries are perfectly capable of competing against low
wage countries due to their higher productivity. What they cannot do is to
compete against the latter in goods in which they have a comparative
disadvantage.
The contention that lower labour standards give poor countries an 'unfair'
advantage begins to look even sillier when we consider the enormous
advantages enjoyed by developed countries in the areas of technology and
capital. Thus, for instance, if we were to poll individuals in New Delhi on
whether the superior access to technology and capital gives developed
countries unfair competitive advantage, almost all of them will say 'yes'; they
will also overwhelmingly support provisions in the WTO that will require
developed countries to share technology with developing countries at low or
no cost. But does that make good economic sense? The very essence of the
gains from trade is that, due to differences in underlying fundamentals,
countries differ in their abilities to produce different products. Developing
countries have a comparative advantage in labour-intensive goods and
developed countries in capital- and technology-intensive goods.
A slightly different twist to the fair trade argument is that trading freely
with countries with lower labour standards may lead to a decline in one’s
own standards. This is the so-called 'race-to-the-bottom' argument. As
commonly made, the argument states that competition for capital and jobs
may lead countries to adopt ever-declining labour standards. Therefore, there
is a need to set labour standards cooperatively. While this theoretical
possibility exists, its empirical relevance depends on two key factors: (1)
responsiveness of capital to labour standards and (2) degree to which
countries compete for capital by lowering labour standards. These are both
empirical questions on which, to date, the proponents of the 'race-to-the-
bottom' argument have provided little evidence. Levinson (1996) looked at
the first of these two questions in the context of environmental standards and
found very weak evidence, at best, in favour of capital mobility in response
to differences in standards. Though no direct evidence is available on the
second question, it is unlikely that countries set labour standards so as to
make them attractive to capital. As already argued earlier with the help of the
Indian example, dominant considerations in setting labour standards are
domestic. Moreover, as Bhagwati (1995) argues, governments typically play

13
the game of attracting capital through tax breaks, land grants at highly
subsidised prices, cheap electricity and so forth.

2.5 Ineffectiveness of Trade Sanctions in Raising Labour Standards

Many developing countries do recognise the need for raising labour

standards. Child labour in India is a case in point. To begin with, poor parents
love their children just as much as rich parents do. They send their children to
work not out of wickedness, but out of sheer economic necessity. Moreover,
there are numerous NGOs in the country working towards alleviating the
child labour problem and the government is under continuous pressure from
them. Finally, there are also laws against the employment of children in
hazardous industries, but their enforcement remains beyond the means and
ability of the government.9 It is unlikely that trade sanctions can significantly
change this reality.
Indeed, there are reasons to believe that trade sanctions will have the
opposite effect to the desired impact. This is evidenced by the experience of
Bangladesh in 1993 when merely the threat of US sanctions led the terrified
owners of garment factories in Dhaka to dismiss from work all children
below the age of 16. According to a recent article by Jeremy Seabrook in the
Financial Times, anecdotal evidence suggests that many of these children
met a fate worse than in the factories, ending up in workshops and factories
not producing for export, or as prostitutes and street vendors.
More broadly, few advocates of trade sanctions against child labour realise
that, world-wide, only five percent of the working children are employed in
export industries. This percentage can be reduced to zero simply by moving
the children to produce similar goods sold domestically and having adults
produce the goods sold in foreign markets. The resulting increase in the costs
of exports will provide some additional protection to the corresponding US
and European industries, but do nothing at all to lower the aggregate level of
child labour.
Likewise, export-processing zones employ a tiny fraction of the labour
force, well below one percent. A stricter enforcement of labour rights will do

8. Developing countries are not alone in the inability to enforce laws in this sphere. Thus,
consider the following statement in de Gray (1994): 'Recent reports about labor conditions in
the US, for example, indicate that there is significant illegal employment of children and that
many such children work in unsafe, dangerous conditions. A spokesman for the US National
Child Labor Committee is reported as saying recently, "We have no compunction in saying
that there are over two million child labor violations each year in the US". As for
enforcement, it is reported that an establishment employing adolescents can anticipate a
federal inspection once every 50 years'. The quotation by de Gray is from Holloway (1993),
which itself appeared in the magazine Scientific American.

14
little for labour rights in general. The bulk of labour force in developing
countries is employed in the informal sector that hardly engages in
international trade. The inevitable conclusion is that the proposed link will do
virtually nothing to improve labour's fate, whether in relation to children or
adults.

2.6 Selectivity in the Developed Country Agenda

A key problem with the manner in which developed countries have pursued
the demand for higher labour standards – regardless of whether it is dealt
with at the WTO or the ILO – is the asymmetry of the proposed agenda.
Independently of whether the objective is the preservation of moral values or
pursuit of fair trade, the issues proposed for negotiation are exclusively those
in which developing countries must give one-way concessions. Thus, only
issues such as child labour and labour rights in export processing zones,
where developing countries will be defendants rather than plaintiffs, are
included. Enforcement against sweatshops in the apparel industry or the
rights of migrant labour which is subject to quasi-slavery conditions in parts
of the industrial world, are not on the proposed agenda.
One suspects that if the issue of child labour is brought into the WTO, the
US will eventually want to set annual targets for its reduction much along the
lines it once sought to expand the Japanese imports from the US. Why then
should we also not have similar targets for unionisation? After all, in the US,
less than 14 percent of labour force is currently unionised and is, thus, able to
bargain collectively. Why should enforcement not be judged by outcomes in
this area? Moreover, unionisation in the US has been handicapped, among
several factors, by legislation on matters such as the right to hire replacement
workers during a strike. Such legislation has blunted the unions' chief weapon
– the ability to strike effectively. Why are the rules that make strikes a more
effective tool of unionisation, not on the proposed agenda for the WTO?
Also missing from the social agenda proposed by developed countries at
the WTO is the subject of union representation on boards of directors that
exists in some European nations. If developing countries are to go along with
conferring on workers, as they should, the right to unionise, mature
democracies such as the US should also be willing to take on additional
obligations that advance the rights of labour in these countries. Yet, no such
move is on the horizon and, in any case, it is an academic issue since, at
present, unions are absent from most US factories in the first place.
Yet another example of protectionist bias in the current labour standards
agenda is the demand for the minimum wage. Why should minimum wage
take priority over employment generation in poor countries where
unemployment is high and employment insurance virtually absent? Will

15
workers, as a whole, be better served by guarantees of minimum wages for
those lucky enough to find jobs, or by greater employment opportunities?
Surely, higher minimum wages in developing countries will provide
industrial countries with greater protection against imports from those
countries. Greater employment opportunities, by contrast, may bring more
labour-intensive goods into industrial countries and hence intensify
competition.
Finally, if the objective is to promote 'social' values, why should this be
limited to the values related to labour? Why not also include issues such as
drug trafficking? Those advocating the trade-labour link often complain that
in the absence of such a link, countries such as India fail to satisfactorily
enforce their own laws relating to child labour. But as Mohan (1996) notes,
one can equally complain about the lax enforcement of the laws against drug
trafficking in the US. Drug trade has been the cause of continued large
overvaluation of the exchange rate and, hence, de-industrialisation in the
neighbouring countries in Latin America. Should these latter countries then,
not have the right to link the access to US exports in their markets with the
effective enforcement of drug laws by the US?

3. The ILO Conventions10

Conceptually, labour standards include all norms and rules that govern
working conditions and industrial relations. But in articulating their demands,
developed countries have focused on some specific standards that they have
gone on to term the '"core" labour standards'. This term gives the impression
that the standards encompassed by it must be so basic that there could hardly
be any disagreement on their desirability. Yet, the matter is not so simple. As
Bhagwati (1995) argues cogently, once we get past slavery, universally
agreed labour standards simply do not exist. For instance, even the issue of
child labour, which at first blush seems straightforward, is quite contentious.
In many countries, the alternative to child labour may be starvation.
Moreover, in the absence of educational facilities, abrupt abolition of child
labour may produce social chaos. Not surprisingly, many countries are
opposed to the abolition of child labour until they reach a level of
development in which the income earned by children is no longer necessary
as a contribution to the family income to avoid starvation, and sufficient
educational facilities are available.
• Freedom of Association and Protection of the Rights to Organize
Convention, 1948 (C87);

9. Unless otherwise noted, the information in this section has been drawn from various pages of
the official ILO web site on the World Wide Web.

16
 • Right to Organize and Collective Bargaining Convention, 1949
(C98);
• Forced Labour Convention, 1930 (C29) and Abolition of Forced
Labour Convention, 1957 (C105);
• Equal Remuneration Convention, 1951 (C100), and Discrimination
(Employment and Occupation) Convention, 1958 (C111); and
• Minimum Age Convention, 1973 (C138) and the Convention on the
Worst Forms of Child Labor, 1999 (C182).

The term '"core" labour standards' has come to mean different things to
different individuals and entities. One particular definition that has been at
the center of the discussion is that used by the ILO. This institution defines
the 'core' standards as consisting of the following five norms (described in
greater detail in the corresponding ILO Conventions):11
Convention 87 stipulates that workers have the right to establish and join
the organisations of their own choosing. These organisations, in turn, have
the right to draw their own constitutions and rules, to elect their
representatives in full freedom, and to organise their administration and
activities. Convention 98 is viewed as essential to making Convention 87
meaningful and is, therefore, frequently grouped with it. Convention 98
provides for protection against anti-union discrimination, for protection of
workers' and employers' organisations against acts of interference by each
other and for measures to promote collective bargaining.
Convention 29 requires the suppression of forced or compulsory labour in
all its forms. It permits certain exceptions such as military service, convict
labour properly supervised, and emergencies such as wars, fires and
earthquakes. Convention 105 prohibits the use of any form of forced or
compulsory labour as a means of political coercion or education, punishment
for the expression of political or ideological views, workforce mobilisation,
labour discipline and punishment for participation in strikes or
discrimination.
Convention 100 provides for equal pay for work of equal value without
discrimination based on sex. Convention 111 calls on the signatory
governments to declare and pursue a national policy designed to promote, by

10. Informal agreement on which of the ILO Conventions are to be regarded as human rights
standards dates at least as far back as 1960. Formal recognition was achieved when the Social
Summit in Copenhagen in 1995 identified six ILO Conventions as essential to ensuring
human rights in the workplace: Nos. 29, 87, 98, 100, 105 and 111. In addition, the United
Nations High Commissioner for Human Rights now includes these conventions on the list of
'International Human Rights Instruments'. The Governing Body of the ILO subsequently
confirmed the addition of the ILO Convention on Minimum Age, No. 138 (1973), in
recognition of the rights of children.

17
methods appropriate to national conditions and practice, equality of
opportunity and treatment in respect of employment and occupation. It calls
for the elimination of any form of job discrimination based on race, colour,
sex, religion, political opinion, national extraction or social origin.
Convention 138 establishes a minimum age for work. It states that children
should not enter the labour market before completion of compulsory
education or before having reached the age of 15. In case of work that is
unhealthy or dangerous, the Convention sets a minimum age of 18. These
provisions are waived in relation to work undertaken in training institutions.
Likewise, light work by children aged from 13 to 15 years may be allowed so
long as it is not prejudicial to their educational activities. Developing
countries may lower the minimum age of employment to 14 years and of
light work to 12 years.
Convention 182, which came into existence as recently as 1999, abolishes
child slavery, child prostitution and pornography, use of children for illicit
activities (e.g., drug trafficking), and any work which is likely to harm the
health, safety or morals of children. Included in the last category is the work
that exposes children to abuse, work with dangerous machines and
equipment, work involving heavy loads, work in unhealthy environments,
work for long hours or during the night and work involving unreasonable
confinement at the employer’s premises.
In the context of the rights of children, it is important to mention the
United Nations (UN) Convention on the Welfare of the Child, 1989. This
Convention is the most comprehensive instrument pertaining to the
protection of children. It is distinguished from the ILO Convention 138 in
that it focuses on 'child welfare' rather than on 'child labour'. The UN
Convention states that children have the right to survival, protection, care,
development and participation. In Article 32, it stipulates that children shall
be protected from economic exploitation and from work likely to be
hazardous or harmful to their health, or to interfere with their education.
Interestingly, the UN Convention does not stipulate a uniform minimum age
for work. While many developing countries have ratified the Convention, the
US has not done so.
Some interesting facts with respect to the ratification record of various
countries may be summarized as follows.12 First, the total membership of the
ILO currently is 175. The number of countries that have ratified the
Conventions are, respectively, 154 for C29, 147 for C105, 132 for C87, 147
for C98, 149 for C100, 145 for C111, C101 for C138, and 41 for C182. There
are 73 countries, 36 in Europe, 15 in the Americas, 5 in Asia and 17 in

11. The ratification status, as described in the following, is taken from the ILO web pages at the
time of writing.

18
Africa, which have ratified all 'core' Conventions excluding C182. Second,
the conventions with respect to child labour remain the least ratified. Some of
the key large, influential developing countries are yet to ratify C138 and
C182 Finally, the US has signed only one of the seven traditional 'core'
Conventions: Convention 105 with respect to the abolition of forced labour.
Thus, the US has not signed the Convention on collective bargaining (No. 98)
or that on child labour (No. 138).
Even if we view the ratification of the Conventions with scepticism, based
on the lack of any requirement for enforcement, non-ratification by some
developed countries, most notably the US, highlights the lack of agreement
on 'core' labour standards. The presumption is that if the practice in a country
is consistent with a particular convention, the country would ratify it.
Applying this logic, one is forced to conclude that the practice and laws in
the US fall short of meeting the standards contained in six of the seven 'core'
ILO Conventions in some form or other. This is most surely the reason
behind the highly selective approach taken by the US in its demands for the
trade-labour link in the WTO.

4. PRESSURE POINTS IN THE CURRENT SYSTEM

The discussion of current pressures can be divided into two categories. In the
first category, we have pressures to introduce the trade-labour link. In the
second category, we have pressures coming from the existing provisions in
the policies of developed countries that permit such a link. Let us discuss
each category in turn.

4.1 The Pressures to Introduce the Trade-Labour Link in the

WTO13

At the governmental level, as opposed to the civil-society level, the pressure

for the trade-labour link at Seattle came almost exclusively from the US. For
instance, according to a 15 November 1999 statement by the Canadian Trade
Minister Pierre Pettigrew, the Canadian objective at Seattle was to 'work for
official observer status at the WTO for the International Labour Organization
(ILO) and urge global compliance with the ILO's two main instruments on
child labour and worker rights'.
According to the document, 'Preparation of the Third WTO Ministerial
Conference: EU Council Conclusions', while the European Union (EU)
Council agreed to strongly support the protection of 'core' labour rights, it

12. The documents on which this section is based can be found in Jenkins (1999).

19
explicitly distanced itself from a trade-labour link. The Council agreed that at
Seattle the EU would advocate three measures for progress on the issue of
trade and labour rights: (1) enhanced cooperation between the WTO and the
ILO and their Secretariats; (2) support to the work of the ILO and its
observership in the WTO; and (3) creation of a Joint ILO/WTO Standing
Working Forum on trade, globalisation and labour issues to promote a better
understanding of the issues involved through a substantive dialogue between
all interested parties (including governments, employers, trade unions and
other relevant international organisations).
The last of these measures may suggest the EU’s acquiescence to the US
position. But the Council went on to explicitly confirm the EU's firm
opposition to any sanctions-based approaches. It also agreed to pursue
international consensus through discussions and negotiations with its
partners. The Council agreed to oppose and reject any initiative to use labour
rights for protectionist purposes and went on to state that the comparative
advantage of countries, particularly low-wage developing countries, must in
no way be put into question.
Thus, at present, the demand for the trade-labour link is coming
exclusively from the US, which has had a long history of making such a
demand. In its 1987 submission to the GATT Council, the US had sought the
establishment of a working party on labour standards with the following
standards on its agenda: (a) freedom of association, (b) freedom to organise
and bargain collectively, (c) freedom from forced and compulsory labour, (d)
minimum age for employment, and (e) measures for setting minimum
standards in respect of conditions of work. Because the GATT Council failed
to reach an agreement, no working party was established. In 1990, the US
renewed its request, but trimmed the list of standards to (a), (b), and (c).
Again, there was no consensus.
Therefore, unlike on environment, the Uruguay Round did not provide for
a committee on labour. The internal US response to this fact was to build the
demand for seeking a future trade-labour link into the Uruguay Round
implementing legislation. Thus, the Uruguay Round Implementing Bill,
Language on the WTO Labor Working Party (H.R.5110), passed by the US
Congress in November 1994, explicitly directs the President to 'seek the
establishment in the GATT 1947, and, upon entry into force of the WTO
Agreement with respect to the US, in the WTO, of a working party to
examine the relationship of internationally recognized worker rights, as
defined in section 502(a)(4) of the Trade Act of 1974, to the articles,
objectives, and related instruments of the GATT 1947 and of the WTO,
respectively'. The objectives of the working party in H.R.5110 are to '(1)
explore the linkage between international trade and internationally recognised
worker rights, as defined in section 502(a)(4) of the Trade Act of 1974,

20
taking into account differences in the level of development among countries;
(2) examine the effects on international trade of the systematic denial of such
rights; (3) consider ways to address such effects; and (4) develop methods to
coordinate the work program of the working party with the International
Labor Organization'.
This legislation formed the basis of the US demand for a WTO working
party at the Singapore Ministerial. As noted in the introduction, the offensive
mounted by the developing country Members of the WTO foiled the US
efforts at Singapore. Notwithstanding the decision there to delegate the
subject of labour standards to the ILO, in November 1999 the US tabled
before the WTO a proposal to create a working group on trade and labour.
In its proposal, the US asked that the working group cover the following
areas:
(1) Trade and Employment: examination of the effects of increased
international trade and investment on levels and composition of countries'
employment.
(2) Trade and Social Protections: examination of the relationship between
increased openness in trade and investment and the scope and the structure of
basic social protections and safety nets in developed and developing
countries.
(3) Trade and Core Labour Standards: examination of the relationship
between economic development, international trade and investment, and the
implementation of 'core' labour standards.
(4) Positive Trade Policy Incentives and Core Labour Standards:
examination of the scope for positive trade policy incentives to promote
implementation of 'core' labour standards.
(5) Trade and Forced or Exploitive Child Labour: examination of the
extent of forced or exploitive child labour in industries engaged in
international trade.
(6) Trade and Derogation from National Labour Standards: examination of
the effects of derogation from national labour standards (including in export
processing zones) on international trade, investment and economic
development.
As it stands, most developing countries oppose the introduction of any
new provisions linking trade and labour. They also see the establishment of a
WTO working group as the first step towards a formal link and, therefore,
stand squarely opposed to it. For instance, it is not clear that the Government
of India will even consider setting up a working group of the ILO and the
United Nations Conference on Trade and Development (UNCTAD) that has
been suggested recently by Bhagwati (1999) as a compromise solution.

21
4.2 The Trade-Labour Link within the Existing Policies

Some developing countries already face the prospect of actual trade policy
action against their exports, based on labour standards. In the US, under the
1984 GSP amendment, the US Government may deny the right to preferential
entry of exports of a beneficiary country that 'has not taken or is not taking
steps to accord internationally recognised worker rights to workers in the
country'. Since 1988, 'denial of workers' rights' has been defined as an unfair
trade practice in section 301 of the US Trade Act of 1974 and may be subject
to action if it harms US economic interests.14
In June 1990, the American Federation of Labor-Congress of Industrial
Organization asked the United States Trade Respresentative for withdrawal
of GSP privileges from Bangladesh on charges of violation of internationally
recognised labour rights.15 The charges included the use of child labour in the
garment industry, violation of the minimum wage legislation, denial of the
right to form trade unions in export processing zones and poor working
conditions. At the time the complaint was filed, 30 percent of Bangladesh's
garment exports to the US were covered by GSP and were subject to either
zero or less than MFN duty. The Export Promotion Board contested the case
for the Bangladesh Government. Fortunately, the US GSP Subcommittee
investigated the case and decided to defer the final decision in April 1991 for
one year.16
The social clause in the EU's new GSP scheme for 1995-1997 also
discriminates against countries that do not apply labour standards contained
in specific ILO Conventions. But the EU decided to suspend, until 1997,
actual implementation of any linkage between GSP and workers' rights, in
order to take into consideration the outcome of discussions in international
negotiations. It was agreed, however, that starting 1 January 1998, GSP
would be tied to effective implementation of 'core' ILO conventions by a
special incentive arrangement and a conditional arrangement. Under the first
arrangement, the EU will grant an additional preference margin to countries
introducing the prohibition of child labour and effective policies for the
protection of workers' rights to organise and bargain collectively. Under the
second arrangement, the EU will have the authority to withdraw temporarily

13. See Raychaudhuri (1996) who, in turn, refers to de Castro (1995) as the source of his
information.
14. Khan (1996) reports that, according to Pakistani news chapters The News, 10 March 1996,
and Dawn, 11 April 1996, Pakistan has also received threats of withdrawal of GSP privileges
from the US. The threat applied to three industries: surgical instruments, sporting goods, and
carpets.
15. Reza et al. (1996), who document this case, do not report the final outcome.

22
some or all of a country's preferential entitlement if the country employs
forced labour or exports goods made by prison labour.17

5. CONCLUSIONS: WHAT ARE THE SOLUTIONS?

A close examination of various arguments in section 2 demonstrated that a
logically consistent case for a link between trade and labour standards does
not exist. Not surprisingly, an editorial in the Financial Times (1999)
preceding the Seattle Ministerial offered the following diagnosis:

The WTO is ill-equipped to tackle these problems, even if its members agreed
to do so. Grafting social policy aims onto an institution designed to dismantle
economic barriers is a recipe for confusion. In any case, poor countries would
reject such a move as disguised protectionism.
Western governments know this. So why are they renewing their demands,
despite a 1996 WTO agreement to leave labour standards to the International
Labour Organization? And why risk splitting the WTO over the issue, just
when unity is needed to launch a new trade round?
Most admit privately that they lack a persuasive intellectual case. But they say
that unless the WTO acknowledges the depth of concern in the west about
labour rights, further trade liberalization will be politically impossible. That
claim looks disingenuous.

The discussion in section 4.1 reveals that even among developed countries,
only the US is seeking a link between trade and labour. The demand for a
WTO working group was tabled by the US alone. Canada did not support this
demand. The EU was willing to go as far as a joint WTO/ILO Forum, but
explicitly ruled out forging a link between labour standards and market
access. Developing countries are outright opposed to a WTO working group
with or without the involvement of the ILO.
Therefore, the solution must be sought in diffusing the pressures for a link
between trade and labour in the US. Given the current mood of civil society,
particularly of labour groups, this is a tough task and may require action on
several fronts. There are no easy options but three suggestions can be made.
First, one of the arguments that civil society groups have made with some
justification recently is that the WTO looks after corporate interests while
ignoring the interests of labour and environment. Two events that have given
substance to this complaint are the inclusion of the Agreement on Trade-
Related Aspects of Intellectual Property Rights (TRIPs), which makes
intellectual property protection a central plank of the WTO, and the OECD

16. Raychaudhuri (1996), who relies on de Castro (1995) for this information, notes that the
current status of this linkage is not known.

23
push for a Multilateral Agreement on Investment (MAI). The civil society
was successful in defeating the latter, but resents the former. It demands that
the WTO must now do for nature and workers what it has already done for
corporate interests. For instance, consider the following excerpt from a
recent article by Bernard (1999) in the Washington Post:

For example, the WTO says its purview does not include social issues – only
trade. So it claims to be powerless to do anything about a repressive regime
selling the products of sweatshops that use child labour. Yet let the same
regime use the same children to produce 'pirated' CDs or fake designer T-
shirts, and the WTO can spring into action with a series of powerful levers to
protect corporate 'intellectual property rights'.

A possible response to this critique is to either take TRIPs out of the WTO
and place it into the World Intellectual Property Rights Organization
(WIPO), or weaken its enforcement provisions. Rather than add to the
existing distortion to achieve equality between corporate and labour interests,
the economically superior option is to reverse the mistake that has been
made.
As far as the environment is concerned, the leadership in the US and
elsewhere in the developed world needs to articulate better the message that
this subject is already under discussion at the WTO. The WTO Committee on
Trade and Environment was created following a Ministerial Decision in
Marrakesh in April 1994. This committee had been given both analytical and
prescriptive functions: to identify the relationships between trade and
environmental measures in order to promote sustainable development, and to
make recommendations on whether any modifications to the provisions of the
multilateral trading system are required. The work of the Committee is in
progress. In addition, environmental concerns have also been aired at the
WTO Dispute Settlement Body with the US environmental laws to achieve
domestic objectives upheld in both the gasoline and shrimp-turtle cases.18
Second, we can strengthen the ILO and begin to promote more effectively
labour standards through this institution. Many developed country WTO
members including the EU, Canada and Australia already favour this
approach. What is needed is to convince the US Government and civil
society groups that this is a feasible approach. Once TRIPs is weakened, this
task may become easier. Bhagwati (1999) has recently proposed the creation
of a joint ILO-UNCTAD Expert Group consisting of trade and labour
scholars. He argues that if the agenda of such a group is symmetric, such that

17. There has been a clear public relations failure in explaining to the public that in each of these
cases, the US law was upheld. The WTO only questioned the implementation that was ruled
unduly discriminatory.

24
it is charged with studying egregious labour practices in both developed and
developing countries, progress towards higher labour standards world-wide is
feasible. At the moment, many developing countries are unwilling to go
along with any kind of working group that brings a trade institution in the
game, while the US wants such a group in the WTO. Therefore, an ILO-
UNCTAD group may be the compromise on which both sides may agree.
Finally, some action can be taken directly by developed countries
themselves. To the extent that they fear capital flight on account of low
labour standards in developing countries, they can require their own
corporations to adhere to the same labour standards abroad as at home. As
Bhagwati (1995) puts it – do in Rome as you do in New York, not as the
Romans do. Such an approach may eliminate a key source of concern on the
part of labour groups.

25
REFERENCES
Bernard, Elaine (1999), 'A New Society will be Heard', Washington Post, 5
December, pp. B1 and B5.
Bhagwati, Jagdish N. (1999), 'Let the Millennium Round Begin in New Delhi',
interview in the Economic Times, 25 January 1999.
Bhagwati, Jagdish N. (1971), 'The Generalized Theory of Distortions and Welfare', in
Jagdish N. Bhagwati, Ronald W. Jones, Robert A. Mundell and Jaroslav Vanek
(eds), Trade , Balance of Payments and Growth: papers in International
Economics in Honor of Charles P. Kindleberger, Amsterdam, The Netherlands:
North-Holland, pp. 69-90.
Bhagwati, Jagdish N. (1997), 'The World Trading System: The New Challenges', in
Arvind Panagariya, M.G. Quibria, and Narhari Rao (eds), pp. 49-79.
Bhagwati, Jagdish N. (1995), 'Trade Liberalization and Fair Trade Demands:
Addressing the Environmental and Labour Standards Issues', World Economy, 18,
745-59.
Bhagwati, Jagdish N. and V.K. Ramaswami (1963), 'Domestic Distortions, Tariffs,
and the Theory of Optimum Subsidy', Journal of Political Economy, 71(1), 44-50.
de Castro, Juan A. (1995), 'Trade and Labor Standards: Using the Wrong Instrument
for the Right Cause', Discussion Chapter No. 99, Geneva, Switzerland: UNCTAD.
de Gray, Rodney (1994), 'The International Trading System and Labor Standards',
Document UNCTAD/MTN/RAS/CB.11, Project RAS/92/034, Geneva,
Switzerland: UNCTAD.
Financial Times 1999, 'Trade and Labor', 13 October 1999.
Holloway, Margarite (1993), 'Hard Times/Occupational Injuries Among Children are
Increasing', Scientific American, 269 (4).
Jenkins, Kristen (1999), 'Dueling Agendas: Documents on Labor and the WTO', New
Economy Information Service, 1 November 1999.
Khan, Ashfaque (1996), 'Study of the Emerging Trading Environment: The Case of
Pakistan', chapter presented to the Emerging Trading Environment and Developing
Asia Conference on Country Studies, 29-30 August, Manila: Asian Development
Bank.
Levinson, Arik (1996), 'Environmental Regulation and Industry Location:
International and Domestic Evidence', in Jagdish N. Bhagwati and Robert E.,
Hudec (eds), Fair Trade and Harmonization: Prerequisites for Free Trade?,
Cambridge, MA, US: MIT Press, pp.429-57.
Mohan, Rakesh (1996), 'Comment on "Observations on International Labor Standards
and Trade" by Alan Krueger', mimeo.
Panagariya, Arvind (1999a), 'WTO: US Demands Threaten Seattle Meeting',
Financial Times, 10 November 1999.
Panagariya, Arvind (1999b), 'The Millennium Round and Developing Countries:
Negotiating Strategies and Areas of Benefits', chapter presented at the Conference
on Developing Countries and the New Millennium Round of Trade Negotiations,
5-6 November 1999, Cambridge, MA, US: Harvard University.
Panagariya, Arvind, M.G. Quibria, and Narhari Rao (eds) (1997), The Global Trading
System and Developing Asia: Challenges and Opportunities, Hong Kong: Oxford
University Press
Raychaudhuri, B. (1996), 'The International Trading System and Social Clauses' in
Asian and Pacific Developing Economies and the First WTO Ministerial
Conference: Issues and Concerns, Studies in International Trade and Investment

26
 22, Economic and Social Commission for Asia and the Pacific, New York: United
Nations.
Reza, Sadrel, M. Ali Rashid, Mustafizur Rahman (1996), 'The Emerging Global
Trading Environment and Developing Asia: Bangladesh Country Chapter', chapter
presented to the Emerging Trading Environment and Developing Asia Conference
on Country Studies, 29-30 August 2000, Manila: Asian Development Bank.
Rodrik, Dani (1997), Has Globalization Gone Too Far?, Washington, DC, US:
Institute for International Economics.

27